HR: 0830h
AN: OS41C-0810    [PDF]
TI: The Ocean's Abyssal Mass Flux Sustained Primarily By the Wind: Vector Correlation of Time Series in Upper and Abyssal Layers
AU: * Hancock, L O
EM: lhancock@worldbank.org
AF: World Bank, MSN H-5-503 1818 H Street, Washington, DC 20433 United States
AB: As Wunsch has recently noted (2002), use of the term "thermohaline circulation" is muddled. The term is used with at least seven inconsistent meanings, among them abyssal circulation, the circulation driven by density and pressure differences in the deep ocean, the global conveyor, and at least four others. The use of a single term for all these concepts can create an impression that an understanding exists whereby in various combinations the seven meanings have been demonstrated to mean the same thing. But that is not the case. A particularly important consequence of the muddle is the way in which abyssal circulation is sometimes taken to be driven mostly or entirely by temperature and density differences, and equivalent to the global conveyor. But in fact the distinction between abyssal and upper-layer circulation has not been measured. To find out whether available data justifies a distinction between the upper-layer and abyssal circulations, this study surveyed velocity time series obtained by deep current meter moorings. Altogether, 114 moorings were identified, drawn from about three dozen experiments worldwide over the period 1973-1996, each of which deployed current meters in both the upper (200$<$z$<$1000) and abyssal (z$>$3750) layers. For each pair of current meters, the Kundu and Crosby measures of vector correlation were estimated, as well as coherences for periods from 10 to 60 days. In the North Atlantic, for example, Kundu vector correlation (50-day window): 0.48 +/- .03\\ Crosby vector correlation (absolute value, 50 day window): 0.46 +/- .07\\ Coherence at 60 days: .36 +/- .07\\ - at 30 days: 0.40 +/- .06\\ - at 10 days: 0.22 +/- .05\\ Most figures for the South Atlantic, Pacific and Southern Oceans are similar. Those obtained in the Indian Ocean or near the Equator are somewhat different. The statistics obtained here are consistent with the work of Wunsch (1997), and tend to confirm Wunsch's result that current velocities at depth are linked with those in the upper layers. Energetics of the circulation that do not take this into account are making an unjustifiable approximation of the physics. These results do not tell us whether time averaged flow on longer time scales might permit distinction of upper layer and abyssal flow components. Some intriguing corollaries do follow. First, the abyssal circulation is not identically the same thing as a global conveyor belt driven by temperature and density differences. Rather, as Wunsch noted (2002), the ocean's mass flux is sustained primarily by the wind. We may add that these wind patterns are about as robust as the temperature differences between equator and pole; this major driver of circulation is not a frail phenomenon. Second, the classical notion of a level of no motion that is also a constant-density surface, an LNM, is inconsistent with the results presented here. Such an LNM would wall off the upper layer circulation from the lower, and as they are not walled off, there can be no such LNM. Third, wind stress is being transmitted down column, presumably to the sea floor.
DE: 1635 Oceans (4203)
DE: 1655 Water cycles (1836)
DE: 4504 Air/sea interactions (0312)
DE: 4532 General circulation
SC: Ocean Sciences [OS]
MN: 2003 Fall Meeting